[linux-dvb] Compro VideoMate T-750 DVB tuner card, any updates?
Rod at Rods.id.au
Thu Feb 21 11:43:06 CET 2008
Hi... new here..
Just wondering if there were any updates with the Compro T-750 dual
Possibly the interesting bit is at the end...
I had a look at "previous posts" but found only that the card was
scanned with some Windoze software??? I don't have that software, but
here is what I know
1/. there is nothing written anywhere that the card uses QT1010 or a
XC2028, seems no one has tried pulling the "Cans" from the PCB, easy
job, when you have the tools ;o)
The QT1010 is the front-end for the DVB stream (top can)
The XC2028 is the front-end for the Analog stream (lower can)...
Most of this, I have E-mailed on the MythTV E-mail list, but
re-verifying it here if anyone missed it..
Here is what I have found on this little card....
The RTC on the card, being a known chip, also linked into the
"Power-SW" wiring from the front panel simplifies the "power saving
feature, also the complexity across differing mobo's and their BIOS
Onboard, it has the following devices.
DVB-in -> QT1010 -> CE6353 -> SAA7135
Analog-in -> XC2028 -> SAA7135
Radio-in -> XC2028 -> SAA7135 (I think it goes thru the XC chip)
Composite-in -> SAA7135
RTC (DS1337)-> SAA7135 (I2C)
EEPROM (HT24LC02)-> SAA7135 (I2C)
CE6353 -> SAA7135 (I2C)
QT1010 -> CE6353 (I2C) I'm sure it does this connection, see below..
I have QT1010, SAA7133/5, XC2028, DS1337 loaded as a modules
The RTC is connected to the I2C buss from the SAA7135, the INT
output is connected thru a custom driver chip (seems like just a custom
driver similar to a ULN2004 driver chip, I doubt its I2C, it wouldn't
buzz out to that chip
QT1010 = 0xA0 DVB Front End (#)
DS1337 = 0xD0 RTC, the Alarm out restarts the computer
HT24LC02 = 0xA0 CMOS 2K 2-wire serial EEPROM (#)
XC2028 = 0x Analog/Radio front End (Difficult to get
address info, as its a BGA)
CE6353 = 0x1E Nordig Unified DVB-T CDFDM Terrestrial
(#) Now, as you notice, the I2C address for the QT1010 and the EEPROM
have the same address (0xA0) I feel (assume, could be wrong) that the
I2C for the QT device is wired to the 2nd port of the CE6353 device
XC2028 (CX8800 = module) or CX88... series modules
CE6353 not sure of the module for this yet... not sure how
programmable it is
Have a PDF of the QT1010, cannot find the linky again
CE6353 looks pin-for-pin compatable for the following (Zarlink Devices)
There is a linky on the Intel web site for cross referencing the CE
Also the I2C address on the chip Doc refers to SADD0:4, in the doc,
it said that "In the current TNIM evaluation application, the 2-wire bus
address is 0001 111 R/ W with the pins connected as
I actually found this rather difficult to understand, but I guess I
sussed it... SADD0:4 is 5 pins that are tied to Vdd or Vss, but the pin
outs state they are N.C. (Non Connected) strange..
For the T-750 the configuration is 0001 111r/w Strange how they
didn't change it, but thats what happens when you follow App notes,...
So, I hope this helps someone getting these little beasties going, I
would love to utilise them ;o)
Ok, that little bit above was posted on the MythTV-users listserv,
no one replied to it ;o(
Ok, now for more information, I havn't found this posted on the
Internet anywhere, so I did some probing myself, if I had the computer I
used to have at work, I could post almost the complete circuit diagram
gained from the PCB, and possibly got myself into some real trouble :P
I don't know what the PRO1A does, but I feel its a port driver, or a
masked ROM, or a Fuse link device...
Someone did mention that the tops of the IC's were damaged, and
difficult to read, if you live in Australia, go out and buy yourself a
bottle of "Eucalyptus Oil" it'll clean those chips up really well for
you, the device is something like a 74ALC74 (or is it ALC174, no matter,
its just a simple chip, driving the switch gear to control the outputs,
nothing really special, I think its driven a bit by the PRO1A device, if
I had that computer (mentioned above) I'd have that figgured out
I2C comms are as above, no more clarity needed I think..
I think the bit that people are having a problem with is the GPIO
connections... I'm not sure how accurate the Windoze scanner is, but
here is what I probed...
Format, is SAA7134 (SAA) -> CE6353 (CE)
SAA Pin:Desig -> CE Pin:Desig
86:GPIO0 -> 49:MDO0
85:GPIO1 -> 50:MDO1
84:GPIO2 -> 51:MDO2
83:GPIO3 -> 52:MDO3
82:GPIO4 -> 53:MDO4
81:GPIO5 -> 56:MDO5
80:GPIO6 -> 57:MDO6
79:GPIO7 -> 58:MDO7
68:GPIO16 -> 48:MOVAL
60:GPIO19 -> 47:MOSTRT
59:GPIO20 -> 61:MOCLK
Next is the GPIO to the PRO1A Device from the SAA
SAA 78:GPIO8 -> PRO1A U5:6 (U5 is the PRO1A Desig)
77:GPIO9 -> U5:7
76:GPIO10 -> U5:8
77:GPIO11 -> U5:9
61:GPIO18 -> U5:12
56:GPIO23 -> U5:13 (or 14) strange, same resistance to either pin from
GPIO23 200-500R (Ohms)
Next, not 100% sure of these being No-Connect... further
investigation (another lunch break)
Ok, I hope this helps get this little cart moving...
More information about the linux-dvb